me

100

Received 18 May 2014

Received in revised form

22 July 2014

Accepted 25 July 2014

Available online 19 August 2014

the vast number of geothermal resources, a great proportion

low-enthalpy geothermal or other thermal energy is Organic

Rankine Cycle (ORC) binary power generator. A noteworthy

example is the 250 kW ORC plant in Chena Hot Springs,

Alaska, which produces electricity from a very low tempera-

ture (74 C) geothermal resource [6].

being impervious to

ad, and high thermal

mal resource). How-

ever, the total capacity of installed geothermal power lags

lled power of PV and

ively. According to

the total geothermal

t 11.2 GW as of May

2012. The average annual growth rate of geothermal power is

about 2%, while that of PV is about 58% during the same period

of 2006e2011 and up to 74% in 2011 (REN 21, 2012) [7].

Li [8] discussed likely factors leading to the low growth rate

of geothermal energy. The main factors include high initial

* Corresponding author.(K. Li).

Available online at www.sciencedirect.com

ScienceDirect

w.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 5 4 9 7e1 5 5 0 5E-mail addresses: likewen@cugb.edu.cn, kewenli@stanford.eduare low temperature (

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 5 4 9 7e1 5 5 0 515498investment, high exploration risk, long payback and con-

struction time, difficulty to assess resource, and difficulty to

modularize. Li [8] also pointed out possible directions to

accelerate the growth of geothermal power. One of the solu-

tions may be the large-scale utilization of TEG technology.

Since 1821, many researches have investigated the appli-

cation of thermoelectric materials. Thacher [9] developed a

thermoelectric power generator using car exhaust heat. The

maximum power output reached 255 W. Kajikawa and Onishi

[10] developed an advanced thermoelectric conversion

exhaust system in a light truck. Maneewan and Chindarksa

[11] investigated the characteristic and performance of TEG

modules for power generation at low temperatures. The unit

achieved a power output of 2.4Wwith a temperature gradient

of approximately 150 C. The conversion efficiency was about3.2%. Niu et al. [12] constructed an experimental thermoelec-

tric generator unit; a comparison of the experimental results

with those from the previously published numerical model

was analyzed. Rinalde et al. [13] developed thermoelectric

generators for electrification of isolated rural homes. Molina

et al. [14,15] designed and improved controlled for thermo-

electric generator used in distributed generation in 2010 and

designed an innovative power conditioning system for the

grid integration of thermoelectric generators, which enables

high power conversion efficiency and reliability of the system.

Hsu [16] developed a low-temperature waste heat system to

utilize the car exhaust heat as well. When the engine rate

boosted to 3500 RPM, 12.4 W of maximum power output was

obtained at an average temperature difference of about 30 C.Shock [17] invented a thermoelectric generator as waste heat

recovery systems in class 8 truck applications and the output

power can reach 255 W (hot and cold side temperature are

about 600 K and 300 K, respectively). There have been few

reports on the TEG systems with a power over 1 kW at low

temperatures.

Numerical modeling of TEG systems has also been inves-

tigated. Esartea et al. [18] analyzed the influence of fluid flow

rate, heat exchanger geometry, fluid properties and inlet

temperatures on the power supplied. Chen et al. [19] assumed

that heat-transfer obeys the linear phenomenological heat-

transfer law and studied the performance of multi-element

thermoelectric-generators. Yamashita [20] developed new

thermal rate equations by taking the temperature de-

pendences of the electrical resistivity and thermal conduc-

tivity of the thermoelectric (TE)materials into the thermal rate

equations on the assumption that they vary linearly with

temperature. Freunek et al. [21] described an analytical model

for thermoelectric generators and found that the influence of

the Peltier heat on the output power was about 40%. Eisenhut

and Bitschi [22] derived an analyticmodel based on convective

heat sources. Liu [23] presented the designs of electricity

generators based on thermoelectric effects using heat re-

sources of small temperature differences. Karabetoglu et al.

[24] reported the approach to characterizing a thermoelectric

generator at low temperatures. Xiao et al. [25] designed a solar

thermoelectric generator using multi-stage thermoelectric

module; the total conversion efficiencywas 10.52%. Suter et al.

[26] established a numerical model for a 1kWe thermoelectricstack for power generation, which may help define the

configuration and operating parameter range that are optimalExperimental

Before the development of 500 W TEG power generator,

module tests, key parameters optimization and structure

design should be done. Therefore, we tested five different

moduleswith different properties in order to find the TEGwith

the maximum output at a specific temperature difference.

Fig. 1 shows the schematic of the module tests. The TEG

module was clamped tightly in between two containers, one

was the hot side with a high temperature and another was the

cold side with a low temperature.

Table 1 lists the properties of each module. Different ma-

terials were used in different modules. Modules 2, 3, and 4

were made of Bi2Te3 while Module 1 and Module 5 were made

of PbTe, which could tolerate higher temperatures than

modulesmade of Bi2Te3. In addition,Module 5 had 241 couples

while the other modules had only 127 couples. Higher cost of

material and many more couples made Module 5 the most

expensive module among those used in this study. On the

other hand, the types and the thickness of the insulation

material used in these TE modules were different.

The heat-conducting medium between the ceramic plate

and liquid block plays a very important role for the TEG sys-

tems. Both thermal conductivity and cost should be taken into

account. In this study, some commercially available heat-

conducting media were used for the comparison tests. The

property data of the heat-conducting media are shown in

Table 2.from a commercial standpoint. Wang et al. [27] presented a

mathematical model of TEG and preliminary analysis of fac-

tors. Kim [28] derived a model describing the interior tem-

perature difference as a function of the load current of a

thermoelectric generator (TEG) and the results showed

approximately 25% of the maximum output power is lost

because of the parasitic thermal resistance of the TE module

used in the experiment.

Thermoelectric generation technology [29], as one entirely

solid-state energy conversion method, can directly transform

thermal energy into electricity by using thermoelectric

transformation materials. A thermoelectric power converter

has nomoving parts, and is compact, quiet, highly reliable and

environmentally friendly. Therefore, the whole system can be

simplified and operated over an extended period of time with

minimal maintenance. In addition, it has a wider choice of

thermal sources. It can utilize both the high- and low-quality

heat to generate electricity. The low-quality heat may not be

utilized effectively by conventional methods such as ORC

technology.

In this study, we built a power generation system using

TEG modules and conducted experiments to measure the

output power at different temperature gradients and other

conditions. We also tested the efficiency of TEG modules

manufactured with different materials. The cost of the power

generator using TEG technologywas estimated and the results

showed that TEG technology was competitive to PV

technology.A 500 W TEG power generator was designed after above

experimental studies. Its schematic is shown in Fig. 2. The hot

of

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 5 4 9 7e1 5 5 0 5 15499Power testing of different TEG modules

We kept the temperature on the hot side at about 200 C byusing a digital thermostat oil bath and used the tap water as

the cooling liquid on the cold side with a temperature of about

20 C. The temperatures of both hot and cold sides weremeasured and the results are shown in Fig. 3. The temperature

was measured using twomicro-thermocouples with very thinwater, which was supplied by the thermostatic water bath

circulator, simulated the geothermal or hot water and pro-

vided heat for the generators. Cooling cyclewas composed of a

water pump and water container. Thermal couples for tem-

perature measurements and displayers were installed at the

inlet and outlet pipes. The values of voltage, current and

power can be collected by the electrical multimeter.

Results and discussion

Fig. 1 e Schematictips. The temperature on the hot side of the modules was

stabilized at about 180 C and that on the cold side at about40 C. The increase in the temperature on the cold side from 20to 40 C was because of the heat conduction from the hot sidethrough the TEG modules. The temperature difference was

stabilized at around 140 C. The results illustrate that the testsystem for thermoelectric power generation was stable.

With the stable temperature difference of 140 C, wemeasured the output power of the five different TEGmodules.

The results are shown in Fig. 4. Three out of the five

Table 1 e Properties of different modules.

Cost(US$)

Materials Size(cm2)

No. ofcouples

Insulator platethickness (mm)

Module 1 10.5 PbTe 16 127 0.8

Module 2 4.5 Bi2Te3 16 127 0.65

Module 3 7.7 Bi2Te3 16 127 0.65

Module 4 3.4 Bi2Te3 16 127 0.6

Module 5 30.8 PbTe 9 241 0.7thermoelectric modules, Modules 2, 3, and 4, generated power

more than 4.5 W.

The power ratio (power generated by each TEG module

divided by the cost) was calculated and the results are shown

in Fig. 5. Obviously, Module 4, with the cost of 3.4 dollar each,

has the highest power-cost ratio.

Note that Module 5 with the highest cost generated less

power and yielded the lowest power-cost ratio at a tempera-

ture difference of about 140 C. The above test results do notimply that Module 5 is valueless because Module 5 was origi-

nally manufactured for operating temperatures as high as

300 C. Thus, Module 5 may work better at high temperaturesthan other modules. Efficiency is the key parameter for power

generation. A theoretical model is frequently used to calculate

the efficiency of TEG. For a single thermoelectric module:

hi P=Q (1)

P I2RL (2)

the module test.Q aTHI KTH TC 12I2RL (3)

Here hi is the efficiency of the thermoelectric module, P is

the output power and Q is the total quantity of heat, a is the

Seebeck coefficient. K is the heat-transfer coefficient, RL is the

load resistance, TH and TC is the temperature of the hot and

cold side respectively.

hi max TH TC

TH$

1 ZT1=2 1

1 ZT1=2 TC=TH

(4)

HereT TH TC=2. ZT is the dimensionless figure ofmeritand its value, ranging from 0.2 to 2, varies with temperature.

Table 2 e Properties of heat-conducting media.

Medium name Thermal conductivity(W/m K)

Cost (US$)

Silicone film 6.0 17.7

Graphite sheets 15.0 2.5

Silicone grease 3.0 24.1

er

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 5 4 9 7e1 5 5 0 515500The efficiency calculated according to the theoretical model

(Eq. (4)) is shown in Fig. 6. The temperature of cold side was

stabilized initially at about 30 C, and then changed to 40 C,

50 C, 55 C, 60 C, 65 C, 70 C, 75 C, 80 C, 85 C, and 90 C

respectively. Based on the current thermoelectric technology,

the value of ZT reaches about 1.0 and the efficiency can hardly

reach 4% when the hot side temperature was around 100 C

(see Fig. 6).

However, the theoretical model (Eq. (4)) for calculating the

efficiency of thermal power generation using TEG may not be

Fig. 2 e The Schematic of the 500 W thvery accurate because the Seebeck coefficient varies with the

temperature or other parameters. Nonetheless we propose a

different approach to estimating energy conversion efficiency

from thermal to electricity. We define this efficiency as the

ratio of the maximum electricity generated from thermal en-

ergy to the total thermal energy per unit volume of hot water.

This efficiency is defined as global efficiency. Using such a

concept, it is easier to estimate how many KW of electricity

could be generated using per unit volume (for example, one

Fig. 3 e Temperatures on the hot and cold sides of the

module.ton) of hotwaterwith a specific temperature difference, which

is a common engineering question of interest. The mathe-

matical models for determining the global efficiency are pre-

sented as follows.

The total heat energy of the hot liquid can be calculated by

the following equation:

Et C$m$DT (5)Here Et is the total thermal energy of the hot side liquid, C is

the specific heat capacity, m is the mass of the hot side liq-

moelectric power generation system.uid,DT is the temperature difference between the cold and hot

sides.

The maximum electricity generated by TEGs can be

calculated:

Wi Pi$t (6)Here Wi is the electric energy generated by TEGs, Pi is the

instantaneous power, t is the time interval at ith step. The

global efficiency of a TEG system can then be computed:

Fig. 4 e The power generated from different modules.

is because hi is the instantaneous energy conservation

efficiency.

Power testing with different parameters

The measurements of output power for Module 4 were con-

ducted when different heat-conducting media were used. The

high temperature on the hot side was provided by a thermo-

static heating station and the hot side temperature was kept

at about 80 C. Tap water with a temperature of 20 C stillserved as the cold side. At the same time, we chose two

different size modules for the test: 40 mm 40 mm and50 mm 50 mm.

The power test results are shown in Fig. 8. The power

generated is proportional to the module's area. In this study,the 40 mm 40 mm size modules were used for the experi-

Fig. 5 e The power-cost ratio of different TEG modules.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 5 4 9 7e1 5 5 0 5 15501hg Xn

i1Wi=Et (7)

Where hg is the global efficiency of a TEG system.

The experimental procedures for measuring global effi-

ciency from thermal to electricity are described briefly as

follow:

(1) Fill the hot side container using water with a tempera-

ture of TH and amass ofm, keep the temperature of cold

side liquid at TC. The total heat energy of the hot liquid

can be calculated using Eq. (5).

(2) Collect the data of the electricity generated by a TEG

system every second until it decreases to zero. The total

electricity generated from the hot liquid can be calcu-

lated using Eq. (6).

(3) The global efficiency of a TEG system can be calculated

using Eq. (7) with the above experimental data.

The efficiency and cumulative electricity output data of

module 4 were measured using the above experimental pro-

cedures by decreasing the hot side temperature from 95 C to30 C and keeping the cold side temperature at 30 C. Theexperimental results are shown in Fig. 7. It is clear that the

growth rate of electricity output slows down when the tem-perature of the hot water on the hot side decrease to 80 C. Theglobal efficiency was about 10%, which is greater than hi. This

Fig. 6 e The efficiency of thermoelectric module at different

ZT and temperature differences.Fig. 7 e Efficiency and cumulative electricity ofments in next section.

According to the results shown in Fig. 8, the modules with

silicon grease generated 1.424 W and 2.146 W for the heat-

conducting media with different size, respectively. The sili-

con grease with a lower thermal conductivity performs better

than other media with a higher thermal conductivity. The

reasonmight be due to air trapped between the two plates and

heat-conducting media. Note that the thermal conductivity of

air is 0.023 W/m K. On the other hand, silicone grease can

adhere at the interface tightly, so it may help the thermo-

electric modules dissipate heat and generate more power.

However, the silicone grease has an obvious disadvantage: it'svolatile and easy to dry, leading to performance reduction.

Silicone film is not recommended because of its high cost.

The graphite sheet may be the most suitable medium due to

its high thermal conductivity and low cost, but the air gap

weakens the heat-conducting performance.

Based on the above experimental results, our approach to

assembling the TEG modules in this study was to adhere the

graphite sheet at themodule's surface using a trace amount ofsilica gel which can tolerate 200 C. This method not onlytakes advantages of the graphite merit, but also avoids the air

gap's negative effect.On the other hand, the thickness of module and heat-

conducting media between two plates may not be the samethermoelectric module (#4) with the hot side temperature

difference.

Fig. 8 e The effect of area and heat-conducting medium

type on output power.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 5 4 9 7e1 5 5 0 515502without any difference, with the increment of modules, the

surface becomes uneven. It leads to a phenomenon which is

similar to the hot spot effect of solar power generation

system that will decrease the efficiency of the thermoelectric

generation system and shorten the lifetime of the modules.

Therefore, this method was proposed to solve this problem

and one layer of 100 modules was used for the trial (another

layer was the control group).

The layer of 100 modules without any treatments could

reach about 130 W at the temperature difference of 64 C.However, the power of the layer treated by our method could

reach 230 W and the temperature difference was just 57 C(cold side was supplied by the tap water which could keep at

about 20 C), shown in the Fig. 9.Experiments about relationship between power and

quantity of modules were conducted. In fact, the power

output is much less than the product of one piece of module

power and quantity on account of the power consumption on

the internal resistance.With the linear fitting results shown in

the Fig. 10, we can predict the needed quantity of modules atany power value; it was useful and meaningful for system

design and economic assessment.

Fig. 9 e Relationship between power and temperature

difference.Power testing of the 500 W power generating system

The 500 W TEG power generator was composed of 96 ther-

moelectric modules (Module 4). 13 liquid blocks (containers)

and some connection boxes were also used in this system.

In order to visually display the experimental results, ten

direct-current bulbs were used to utilize the output energy. As

shown in Fig. 11, ten 15 W bulbs (direct current) were lit up

with 100 C water on the hot side and 20 C tap water on thecold side of the system.In order to study the effect of the pattern to connect TEG

modules on the power output, 10 different number (1, 2, 3, 4, 5,

10, 20, 50, 70, 100 modules) of modules were used for the

power testing. All of the experiments were conducted under

the temperature difference about 80 C (hot side was about100 C and the cold side was about 20 C supplied by tapwater), the linear fitting results is good. When the module

number is 100, the power output is about 165 W. Based on the

results, about 600 modules are needed for the 1 kW thermo-

electric generator.

Fig. 10 e Relationship between power and quantity of

modules. (Temperature difference was about 80 C).The thermoelectric power generation systemwas operated

using 100 C hot water and 20 C tap water. The 100 C hotwater was supplied by a thermostatic water bath. The power

generated by the system at a temperature difference of about

80 C is shown in Fig. 12. The power was stabilized at around160 W.

The power output was measured at different temperature

differences using the TEG power generator. The purpose was

to establish the relationship between power output and tem-

perature difference. The experimental results are shown in

Fig. 13.

One can see that the power output increases with the in-

crease in temperature differences almost linearly. We can

estimate the power output approximately at a specific tem-

perature difference. For example, a power output of 500Wwill

be reached at a temperature difference of about 200 C. Notethat the slope of the power curve shown in Fig. 13 increases

with the increase in temperature difference. The relationship

between power output and temperature difference looks like

exponential, which is of great significance.

Comparison of cost with PV and wind systems

Cost is a great challenge for almost all of the renewable energy

technologies. Based on the experimental results and the data

about PV and wind system from publications [30,31], we esti-

mated the total cost of the DHE power generator and found

that the cost is close to the PV's cost in terms of unit installed

Fig. 11 e (a) The thermoelectric power generation system lighting up ten 15 W bulbs before assembled; (b) The power

e 2013 exhibition at Shanghai International Expo Center after

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 5 4 9 7e1 5 5 0 5 15503The same experimental method as that applied for single

modules was used for the efficiency determination of the

500 W power generation system. As shown in Fig. 14, the

global efficiency was about 9% with the temperature variation

from 100 C to 30 C of the hot side. The slope gradually de-creases to zero, and the temperature of hot side from 100 C to80 C was recommended for power generation based on theresults in Fig. 14.

On the other hand, temperature differences of inlet and

outlet of hot side also plays very important role on the effi-

ciency. Both inlet and outlet temperature variation of hot side

was collected, inlet temperature was changed from 45 C to85 C and the outlet was from 35 C to 75 C. The temperatureon the cold side was kept at 35 C.

The data shown in Fig. 15 are instantaneous efficiency

which increases with the increase in temperature at the inlet

and the decrease in temperature at the outlet on the hot side.

The instantaneous efficiency of the TEG system could reach

about 4.5% at the inlet temperature of about 95 C and theoutlet temperature of 75 C. Importantly the instantaneous

generation system operated to start the LED display during th

assembled.efficiency increases with the inlet temperature exponentially,

and has a stronger dependency on the outlet temperature.

Note that the results shown in Fig. 15 are similar to the results

reported by Sulter et al. (2012).

Fig. 12 e Power generated from the TEG power system vs.

time.Fig. 13 e Relationship between power and temperature

difference.

Fig. 14 e Global efficiency and cumulative electricity of

geothermal power generation with the hot side

temperature variation.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 5 4 9 7e1 5 5 0 515504Fig. 15 e Efficiency of power generation at different inlet

and outlet temperatures of hot side.power. The cost data for solar PV and wind energy are shown

in Fig. 16. Also shown in Fig. 16 are the cost data estimated by

considering the capacity factor for PV and DHE. A capacity

factor of 14% for PV and 90% of DHE (or TEG technology for

thermal energy) were chosen. The cost of the TEG system

developed in this study was much lower than those of PV and

wind power systems in terms of equivalent energy generated

after considering the capacity factor.

Conclusions

According to the current study, the following preliminary

conclusions may be drawn:

(1) A power generator has been built using TEG modules

and tested. The power could reach about 500 W (pre-

dicted using experimental data) with a temperature

difference of about 200 C between hot and cold sides.An output power of over 160 W has been generated

under a temperature difference of 80 C (hot side tem-perature was about 100 C and the cold side was 20 C).

(2) The instantaneous efficiency of the TEG system reached

4.5% at an inlet temperature of about 95 C on the hotside and a temperature of 30 C on the cold side. Thisefficiency increases exponentially with the inlet tem-

perature when outlet temperature was kept constant.

Fig. 16 e The cost comparison of wind, solar,

thermoelectric power generation.(3) The cost of the TEG system developed in this study was

lower than those of PV and wind power systems in

terms of equivalent energy generated (considering the

capacity factor).

r e f e r e n c e s

[1] WEA. World energy assessment report. Energy and thechallenge of sustainability; 2000. p. 500.

[2] Erdlac Jr RJ, Armour L, Lee R, Snyder S, Sorensen M,Matteucci M, et al. Ongoing resource assessment ofgeothermal energy from sedimentary basins in Texas. In:Proceedings of 32nd workshop on Geothermal ReservoirEngineering. Stanford, California: Stanford University; 2007.

[3] Li K, Zhang L, Ma Q, Liu M, Ma J, Dong F. Low temperaturegeothermal resources at Huabei Oilfield, China. GRC Trans2007:31.

[4] Xin S, Liang H, Hu B, Li K. Electric power generation from lowtemperature co-produced geothermal resources at HuabeiOilfield. In: Thirty-seventh workshop on GeothermalReservoir Engineering Stanford University; 2012.

[5] Bennett K, Horne RN, Li K. Power generation potential fromcoproduced fluids in the Los Angeles Basin. GRC Trans2011;35:85e9.

[6] Erkan K, Holdman G, Blackwell D, Benoit W. Thermalcharacteristics of the Chena Hot Springs Alaska geothermalsystem. In: Proceedings, thirty-second workshop onGeothermal Reservoir Engineering Stanford University,Stanford, California; 2007.

[7] Renewables: global status report, REN21 (renewable energypolicy network for the 21st century); 2012. Available at: www.ren21.net.

[8] Li K. Comparison of geothermal with solar and wind powergeneration systems. In: Proceedings, 38th workshop onGeothermal Reservoir Engineering Stanford University,Stanford, California; 2013.

[9] Thacher E, Helenbrook B, Karri M, Richter CJ. Testing of anautomobile exhaust thermoelectric generator in a light truck.Proc Inst Mech Eng Part D J Automob Eng 2007:95e107.

[10] Kajikawa T, Onishi T. Development for advancedthermoelectric conversion exhaust thermoelectric generatorin a light truck. Proc Inst Mech Eng Part D J Automob Eng2007;221:95e107.

[11] Maneewan S, Chindaruksa S. Thermoelectric powergeneration system using waste heat from biomass drying. JElectron Mater 2009;38(7):974e80.

[12] Niu X, Yu J, Wang S. Experimental study on low-temperaturewaste heat thermoelectric generator. J Power Sources2009;188:621e6.

[13] Rinalde GF, Juanico LE, Taglialavore E, Gortari S, Molina MG.Development of thermoelectric generators for electrificationof isolated rural homes. Int J Hydrogen Energy2010;35(11):5818e22.

[14] Molina MG, Juanico LE, Rinalde GF, Taglialavore E, Gortari S.Design of improved controller for thermoelectric generatorused in distributed generation. Int J Hydrogen Energy2010;35(11):5968e73.

[15] Molina MG, Juanico LE, Rinalde GF. Design of innovativepower conditioning system for the grid integration ofthermoelectric generators. Int J Hydrogen Energy2012;37(13):10057e63.

[16] Hsu C, Huang G, Chu H, Yu B, Yao D. Experiments andsimulations on low-temperature waste heat harvesting

system by thermoelectric power generators. Appl Energy2011;88(4):1291e7.

[17] Schock H, Brereton G, Case E, D'Angelo J, Hogan T, Lyle M,Maloney R, et al. Prospects for implementation ofthermoelectric generators as waste heat recovery systems inclass 8 truck applications. J Energy Resour Technol2013;135(2):022001.

[18] Esartea J, Minb G, Roweb DM. Modeling heat exchangersfor thermoelectric generators. J Power Sources2001;93:72e6.

[19] Chen LG, Sun FR, Wu C. Thermoelectric-generator withlinear phenomenological heat-transfer law. Appl Energy2005;81:358e64.

[20] Yamashita O. Effect of linear temperature dependence ofthermoelectric properties on energy conversion efficiency.Energy Convers Manag 2008;49:3136e69.

[21] Freunek M, Mueller M, Ungan T, Walker W, Reindl LM. Newphysical model for thermoelectric generators. J ElectronMater 2009;38:1214e20.

[22] Eisenhut C, Bitschi A. Thermoelectric conversion systembased on geothermal and solar heat. In: 25th InternationalConference on thermoelectrics; 2006. p.510e5.

[23] Liu L. Large-scale ocean-based or geothermal power plantsby thermoelectric effects; 2012. www.arXiv.org.

[24] Karabetoglu S, Sisman A, Fatih Ozturk Z, Sahin T.Characterization of a thermoelectric generator at lowtemperatures. Energy Convers Manag 2012;62:47e50.

[25] Xiao J, Yang T, Li P, Zhai P, Zhang Q. Thermal design andmanagement for performance optimization of solarthermoelectric generator. Appl Energy 2012;93:33e8.

[26] Suter C, Jovanovic Z, Steinfeld AA. 1 kW thermoelectric stackfor geothermal power generationemodeling and geometricaloptimization. Appl Energy 2012;99:379e85.

[27] Wang Y, Dai C, Wang S. Theoretical analysis of athermoelectric generator using exhaust gas of vehicles asheat source. Appl Energy 2013;112:1171e80.

[28] Kim S. Analysis and modeling of effective temperaturedifferences and electrical parameters of thermoelectricgenerators. Appl Energy 2012;102:1458e63.

[29] Rowe DM. Review: thermoelectric waste heat recovery as arenewable energy source. Int J Innov Energy Syst Power2006;1:13e23.

[30] CWEA. China wind power outlook 2012; 2012. Available at:www.greenpeace.org.

[31] Qian BZ. Review of photovoltaic applications and installationcosts. Sol Energy 2012;16:49e50.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 9 ( 2 0 1 4 ) 1 5 4 9 7e1 5 5 0 5 15505

A 500 W low-temperature thermoelectric generator: Design and experimental studyIntroductionExperimentalResults and discussionPower testing of different TEG modulesPower testing with different parametersPower testing of the 500 W power generating systemComparison of cost with PV and wind systems

ConclusionsReferences